by André Freitas, Edward Curry, Seán O'Riain
Abstract:
The process of searching and understanding existing vocabularies (terminological artifacts) on the Linked Data Web is an intrinsic activity to the consumption and production of Linked Data. Data consumers trying to find and understand the vocabularies behind datasets in order to query them, or data producers searching for existing resources to describe their data, face the challenge of semantically searching existing concepts in vocabularies. Traditional search mechanisms do not address the level of semantic matching necessary to match users' information needs to vocabulary elements, bringing an additional barrier to the consumption and production of Linked Data on the Web. This work describes a terminological search mechanism which uses a distributional semantic model to provide a best-effort semantic matching solution. The distributional semantic model leverages the semantic information present in large volumes of unstructured text to improve the semantic matching capabilities of the search process. A quantitative evaluation of the quality of the search results shows that the approach provides an effective semantic matching mechanism for terminological search.
Reference:
André Freitas, Edward Curry, Seán O'Riain, "A distributional approach for terminological semantic search on the Linked Data Web", In Proceedings of the 27th Annual ACM Symposium on Applied Computing - SAC '12, ACM Press, New York, New York, USA, pp. 384-392, 2012. [slides]
Bibtex Entry:
@inproceedings{Freitas2012,
abstract = {The process of searching and understanding existing vocabularies (terminological artifacts) on the Linked Data Web is an intrinsic activity to the consumption and production of Linked Data. Data consumers trying to find and understand the vocabularies behind datasets in order to query them, or data producers searching for existing resources to describe their data, face the challenge of semantically searching existing concepts in vocabularies. Traditional search mechanisms do not address the level of semantic matching necessary to match users' information needs to vocabulary elements, bringing an additional barrier to the consumption and production of Linked Data on the Web. This work describes a terminological search mechanism which uses a distributional semantic model to provide a best-effort semantic matching solution. The distributional semantic model leverages the semantic information present in large volumes of unstructured text to improve the semantic matching capabilities of the search process. A quantitative evaluation of the quality of the search results shows that the approach provides an effective semantic matching mechanism for terminological search.},
address = {New York, New York, USA},
annote = {<a href="http://www.slideshare.net/andrenfreitas/freitas-acm-sac2012">[slides]</a>},
author = {Freitas, Andr{\'{e}} and Curry, Edward and O'Riain, Se{\'{a}}n},
booktitle = {Proceedings of the 27th Annual ACM Symposium on Applied Computing - SAC '12},
doi = {10.1145/2245276.2245350},
file = {:Users/ed/Library/Application Support/Mendeley Desktop/Downloaded/Freitas, Curry, O'Riain - 2012 - A distributional approach for terminological semantic search on the Linked Data Web.pdf:pdf},
isbn = {9781450308571},
keywords = {Distributional,Explicit Semantic Analysis,Linked Data.,Semantics,Terminological Search,Treo,Vocabulary Search},
mendeley-tags = {Treo},
pages = {384--392},
publisher = {ACM Press},
title = {{A distributional approach for terminological semantic search on the Linked Data Web}},
url = {http://www.edwardcurry.org/publications/Freitas_SAC_2012.pdf},
year = {2012}
}